Proposed is a population-based longitudinal study to characterize the relationship between levels of physical activity and subsequent risk of depression in the elderly. The study will also determine the influence on these factors of the environment in which an older adult lives, and evaluate the impact of a major environmental intervention designed to improve walkability. Physical activity and mental health are fundamental to well-being in the elderly and are closely related in cross sectional research. There is suggestive, but limited, evidence from clinical studies that activity may be protective against symptoms of depression. However, whether this is true in the general population is far from clear. There is also mounting, but limited, evidence to suggest that environmental factors are an important influence on both levels of physical activity and depression. Neighborhood characteristics may be particularly important in later life as older adults are likely to spend more time in their neighborhood of residence, may be more affected by factors such as physical access, and more vulnerable to threats to their safety. Combined, this evidence suggests a complex interaction in older people between physical activity, the environment, and their functional and mental health. The proposed study will examine these issues in a diverse cohort of 3000 residents of New York City (NYC) aged between 65 and 75 years of age recruited by spatially stratified randomized telephone sampling. Participants will be interviewed at recruitment and at two further waves approximately 12 months and 36 months later, using validated instruments to ascertain symptoms of depression, physical activity levels and key individual level covariates. Participants will also use pedometers to objectively measure their activity levels. Neighborhood characteristics will be objectively assessed using a comprehensive geospatial database already developed by the Investigators from a range of secondary data sources, visual assessment and population survey. Patterns of depression and levels of physical activity over the study period will be characterized, and multilevel methods used to examine the relationships between symptoms of depression and physical activity levels, and the influence on both of neighborhood and individual level determinants. Finally, we will test the potential benefits of environmental intervention on these outcomes by evaluating the impact of a major program to improve pedestrian safety for older adults to be implemented in 10 neighborhoods within the study area between study waves 1 and 2. This intervention will be undertaken by study partners the NYC Department of Transportation and has been described as the largest intervention of its kind in the nation's history. We will prospectively evaluate the impact of this initiative in an oversample of 2000 older residents of intervention and comparable control sites. Outcomes will include participant perceptions, behaviors and symptoms of depression. The findings of this innovative research will guide future investment and evidence based policies to help people remain healthy and independent as they age.